
Catch broken guidance connections
A guide can look complete while pointing to a missing contract or a skill that no longer exists. Those broken connections waste time exactly when an agent needs a reliable next step.
Wildo audits the relationships between source topics, references, skill identities and methods. Findings identify the broken connection so it can be repaired at its source.
Example — Rename a contract without stranding its guide
A maintainer changes a named source topic. The audit can identify a reference that still names the old topic, allowing the registry, source and guide to be updated together.
For engineers
Run the maintainer audit
These commands belong in the framework workspace:
# Report failures and warnings under the normal policy.
wildo fwk audit
# Require warnings to be resolved as well.
wildo fwk audit --strict
# Inspect source-topic relationships to their guidance consumers.
wildo fwk graph
Normal success requires no failures. Strict mode additionally requires no warnings. A warning is therefore visible in a successful normal run; inspect the report rather than treating the exit status as the entire result.
Read findings by the relationship they check
| Connection | What can go wrong |
|---|---|
| Source span to registered topic | Missing, unclosed or unregistered anchor |
| Reference to source topic | A placeholder names something the compiler cannot resolve |
| Skill to index | Guidance exists but its routing information has drifted |
| Playbook to skill | A method requests a guide that is absent |
| Guidance artifact shape | A file or declared identity violates the supported format |
The checks cover named relationships and supported patterns. New contracts still need meaningful authored explanations, and an installed agent host still needs to load the delivered entry points.
Keep integrity separate from execution
A reference that resolves may contain an example with an invalid API use. The example-drift checks and owning package tests cover different questions from this audit. Likewise, a green audit cannot prove an application interaction works.
Use the finding’s source location to fix the connection, regenerate the payload, and verify the installed guidance where delivery changed. For a behavior example, also run the check that exercises that behavior.